基于小面元的多源遥感影像高精度配准方法（High accuracy registration method for multi-source remote sensing image based on small surface element）.doc

基于小面元的多源遥感影像高精度配准方法（High accuracy registration method for multi-source remote sensing image based on small surface element） Article no. : 1009-427x (2003) 02-0124-05 High accuracy registration method for multi-source remote sensing image based on small surface element Xing shuai, tan bing, li jiansheng, xu qing, geng xun (school of surveying and mapping, university of information engineering, zhengzhou, henan, 450052) Abstract: based on a few high-precision matching methods, this paper introduces the idea of small surface element differential correction, and designs a kind of accurate image comparison method. real The results show that this method has a strong adaptability to multiple remote sensing images, and the relative accuracy is high, especially for the obvious or multi-source image in the mountainous area High precision registration. Key words: multi-source remote sensing image; High accuracy registration; Matching; Small face correction Middle image classification no. : TP751 literature identification code: A Multi-source remote sensing satellite images refer to different electromagnetic waves. Different time phases, different incident angles, different imaging mechanisms, no The spatial resolution of the same region is different from that of the same area. But you can complement each other [1]. We are just making use of its letter The characteristics of mutual complementarity and cooperation, through certain treatment methods, such as melting The combination, classification, change detection, identification, etc., extract useful or I These methods are applied to the information of interest Premise and foundation. Especially high accuracy registration has become a shadow Like the necessary conditions for obtaining accurate information, meanwhile, high precision image matching Time change detection, stereo matching, motion analysis and graph Basic [2] applications such as sequence analysis. 1. Overview of image registration algorithm At present, the method o